Re: [Gimp-user] image alignment/registration

2017-03-29 Thread Casey Connor
So, by looking at the help text, it seems that you used all the options suited to focus stacking. At least in the help file I'm looking at, I think the only option "for" focus stacking is -m ? For lenses with focus breathing (i.e. almost all lenses) if you are changing the focus for each

Re: [Gimp-user] image alignment/registration

2017-03-29 Thread Casey Connor
D'oh, the -d option! And -x -y -z sound promising as well. Not clear what -i does -- doesn't it always optimize the center shift? Or does "center shift" mean something more complex than simple x/y translation? Thanks, Pat! Pays to RTFM. The hugin toolkit continues to amaze me. That generates

Re: [Gimp-user] image alignment/registration

2017-03-27 Thread Pat David
Could I have another link to the images to test something? On Mon, Mar 27, 2017 at 12:59 PM oneaty wrote: > >One last thought: if you are going to remove the barrel distortion in > >those images, you might try removing it before doing the image > >alignment... that might

Re: [Gimp-user] image alignment/registration

2017-03-27 Thread Casey Connor
One last thought: if you are going to remove the barrel distortion in those images, you might try removing it before doing the image alignment... that might allow align_image_stack to work better... -c ___ gimp-user-list mailing list List address:

Re: [Gimp-user] image alignment/registration

2017-03-27 Thread Casey Connor
Then I take the resulting output.tif file, open it in gimp, change the precision to 8 or 16bit, and remove the alpha channel (right-click the layer). This is the result of that process: http://caseyconnor.org/pub/image/aligned_and_medianed.tif It certainly looks very good! However, it's not

Re: [Gimp-user] image alignment/registration

2017-03-26 Thread Casey Connor
Ok -- I believe PS is beating align_image_stack because the picture seems to have been taken with a very wide lens (i.e. short focal length) from close up... as a result, there is a lot of barrel distortion, and in addition there is a lot of camera movement between frames, so the result is

Re: [Gimp-user] image alignment/registration

2017-03-25 Thread Casey Connor
I tried align_image_stack -a align -g 10 -t 5 *.tif Did you try with a lower-than-default -g? E.g. "-g 3"? If you want to upload the base .tif files somewhere, I can try to align them, so we can see if there is some strange difference between align_image_stack on our machines. -c

Re: [Gimp-user] image alignment/registration

2017-03-24 Thread Casey Connor
Ah, ok. So what exactly goes wrong: one or two of the images is off-kilter? If you specify -g, try using something besides the default, as I'm not sure that using the default will change the behavior. (Meaning, not specifying -g probably results in the same thing as just "-g" -- the

Re: [Gimp-user] image alignment/registration

2017-03-24 Thread Casey Connor
So, backing up a step: when you say align_image_stack doesn't work, what do you mean exactly? I assumed you meant that it did a poor job aligning the images, but I don't think the command lines you list there would even work to start the program, so I'm wondering now if you meant that it

Re: [Gimp-user] image alignment/registration

2017-03-24 Thread Casey Connor
Did you try the -g option with align_image_stack? Or -t? I align hand-held images with align_image_stack and it works fine, even without those options... e.g. this one: https://www.flickr.com/photos/lacinato/32628735771/in/dateposted-public/ -c On 03/24/2017 12:58 PM, oneaty wrote: Per

Re: [Gimp-user] image alignment/registration

2017-02-05 Thread Pat David
Align_image_stack is usually the best option other than possibly attempting it manually directly in Hugin. On Sat, Feb 4, 2017 at 8:57 PM Casey Connor wrote: > Hi - I was curious what the best overall image alignment plugin/method > in GIMP (or out of GIMP) was